1. Introduction

1.1 Agricultural production zones
1.2 Water supplies
1.3 Livestock production systems

1.1 Agricultural production zones

The agricultural production zones of tropical Africa in which livestock are kept can be defined in different ways. Here we shall use a classification of four categories:

Arid zone: Rainfed crop cultivation is uneconomic, given current technology and price ratios, because of very low and unreliable rainfall. Semi-arid zone: Some rainfed cultivation is economic, but due to low and unreliable rainfall crop yields are generally poor and highly variable.

Highland areas of plentiful rainfall: These are mainly found in East Africa, in Ethiopia, Kenya and Tanzania. Rainfall is usually sufficient and temperatures are moderate enough for temperate zone crops to be grown.

Low-altitude areas of high rainfall: These include what are normally classified as humid and sub humid areas. High temperatures discourage the growing of temperate zone crops but rainfall is high enough to allow profitable crop production.

Different criteria and definitions can lead to somewhat different estimates of the size and importance, in terms of livestock, of the different zones. Table 1 indicates the proportions that the populations of livestock and of humans and the surface areas of these zones constitute of the totals for tropical Africa¹. The figures for the low-altitude/high-rainfall zone combine those of what are normally called the humid and subhumid zones: this zone includes much of the land area that tsetse fly infestation renders unsuitable for most livestock.

1 The definition of the zones used in drawing up Table 1 was done in terms of the length of the plant growing period during the year. This is not identical with the classification used in the text above but yields broadly similar results.

Table 1. Relative area and populations of different zones of tropical Africa (all figures as % of total for tropical Africa).

Source: Jahnke (1982).
^a TLU = tropical livestock unit of 250 kg liveweight.

1.2 Water supplies

Water supplies for livestock can also be classified in a number of different ways. The most common classifications are by source of water (e.g. rivers, lakes, dams, wells), by method of extraction (e.g. direct access by animals, rope and bucket, motor-driven pump), by ownership (government, community, private) or by period of use (wet or dry season, ephemeral or permanent). The Appendix to this report sets out and recommends a classification of water points in terms of sources of water. The terms defined in the Appendix are used in the text without further explanation.

Each of the different categories of water supply, defined in terms of water source, can be found in each zone. Neither the relative number of each kind of source nor the volume of water or number of livestock supplied - even if these figures were available which they usually are not - would fully reflect each category's relative importance in each zone. Relative importance is often primarily a function of location, especially in relation to feed resources, and of reliability at critical periods in particularly dry years. However, for purposes of illustration Table 2 shows the water resources which livestock owners used to water their livestock in the dry season in the mid-1970s in the livestock zone of Tanzania. Different regions are ranked by their dryness. There is a tendency for livestock owners in the less dry regions to make relatively more use of streams and rivers and for those in the relatively drier regions to make more use of wells, boreholes and piped water supplies. The latter are all structures in which resources have had to be invested to provide water that would otherwise be absent. There is only a weak (negative) correlation between use of dams and relative dryness, probably reflecting the fact that although there may be a tendency for relatively more of these dams to be constructed in drier regions, dams often dry up before the height of the dry season. The pattern demonstrated by these regions of Tanzania probably reflects fairly well the situation in tropical Africa as a whole. Individual areas in Africa, however, are likely to show marked deviations from the average in response to local factors.

In the drier (arid and semi-arid) areas the importance of water supplies lies in their reliability at times of acute shortage, in their location in relation to supplies of forage and topographical features such as steep escarpments, and in the method by which water is extracted from its source and distributed to livestock. The reliability of supply determines whether herdsmen can keep livestock at all in an area where, in the event of breakdown or drying up, alternative supplies are not available within a 2 or 3 days' trek, and deaths from thirst therefore may occur. The location of water supplies in relation to topographical features and to the available forage determines how much time, energy and body water livestock have to expend in travelling to the water point. The method of extraction determines how much human labour and time is spent in the actual process of watering.

Table 2. The relative use of different kinds of water source by livestock in the dry season in Tanzania (all figures in % of livestock owners claiming this as main source).

Source: Texas A and M University (1976, p.52).

^a) Dryness is measured in terms of the proportion of a region's surface area with 80% probability of annual rainfall exceeding 500 and 750 mm.

^b) Totals may not be 100% due to rounding up of figures.

^c) For seven pairs the values of the coefficient required for significance are 0.56 and 0.75 at the 5% and 1% significance levels respectively.

In high rainfall areas at both low and high altitudes the main function of water supplies and water development is to distribute the already available water more evenly. This is done in order to enable high-yielding dairy animals to be watered several times a day and to reduce the risk of high-value (especially exotic) livestock being exposed to disease infection. For example, in areas where East Coast Fever is a major problem, cattle which have been subject to a rigorous tick-control regime should not have to run the risk of picking up infected ticks from other livestock while trekking across other farms to water. A water supply needs to be made available on-farm. To a lesser extent water development may be necessary in these areas to prevent soil erosion caused by thousands of animals converging on a single water source. In such high-rainfall areas it may, therefore, be appropriate to distribute water by piping it to small farm holdings or to small paddocks. In drier areas disease and erosion are usually of less importance, partly because exotic livestock are not so well adapted to the local environment and therefore less likely to be present, and partly because the lower economic value of land makes soil erosion from the tracks of cattle of any kind of less economic consequence.

1.3 Livestock production systems

In this study, six livestock production systems are distinguished and discussed. These six categories do not cover the spectrum of livestock production in Africa. Attention is focused only on those systems in which water supply and management seem to be particularly critical, and a number of important systems, e.g. all those in the high-rainfall/low-altitude zone, are not considered. Categorising particular cases into systems is a convenient way of sorting out a mass of detail drawn from different places, societies and times so that regular patterns of behaviour can be identified and analysed. The categorisation employed uses individual livestock owners, rather than areas or economic relations, as the units of which systems are composed. Hence in the same geo graphical area two or more 'livestock production systems' can coexist (for an example see Horowitz, 1972), and individual livestock owners can switch, over time, from one system to another. The evidence available about how systems work is usually drawn from a particular place, a particular society (often an ethnic group) and a particular moment in time. In using evidence which describes specific cases to illustrate systems this study uses the present tense, except where it is positively known how the situation has changed since the time when the evidence was collected.

Three of the six categories of production systems mentioned above are described in some detail as follows, as examples from them are discussed at some length later in the report.

1.3.1 Nomadic pastoralism in arid and semi-arid areas

Several species of livestock are normally kept in these systems. Crop cultivation plays, at most, only a minor role, although earnings from wage employment outside the pastoral sector or from trading or caravaneering may contribute substantially to income. Herds and families are extremely mobile; wet-season movements are opportunistic although in the dry season there is a tendency to return to the same water sources each year. Arid areas may be exploited during one season and semi-arid areas during another².

2 Examples are the Somali of the Haud in Ethiopia (Cossins, 1971a), of northern Somalia (Lewis, 1961) and of northern Kenya (Chambers, 1969), the Kababish Arabs of Kordofan Province in Sudan (Asad, 1970), some Tuareg of Mali (Swift, 1979) and Niger (Bernus, 1981), end the WoDaaBe Fulani of Niger (Horowitz, 1972).

1.3.2 Seminomadic pastoralism in semi-arid areas

In such systems, there is a tendency to concentrate on one species of livestock. Movement between, for example, wet and dry-season pastures takes place but it is on a smaller scale than in the nomadic system, and some members of the family, e.g. wives with children, may stay in one place throughout a year. A substantial proportion of families may cultivate some crops, although not necessarily every year. Such crop cultivation may be of equal importance as animal husbandry in terms of sales or subsistence consumption, but animal husbandry is usually accorded primacy in terms of prestige, the attention of the household head, and the allocation of adult male labour. Dwellings may be of either a mobile (tent) or a fixed (house) type³.

3 Examples are the Karimojong of Uganda (Dyson-Hudson, 1972) and some Tuareg of Niger (Bernus, 1981).

1.3.3 Livestock husbandry in mixed farming in semi-arid areas

In these systems, crop cultivation takes precedence over livestock husbandry both in terms of the proportion of income (including subsistence) derived from it and in the allocation of family labour, including the labour of adult males. There may be some movement of livestock between seasonal pastures but, except where livestock are entrusted to the care of pastoralists who are not members of the family (and who may be from another ethnic group), such movements are usually on a small scale and may principally be undertaken in order to keep livestock away from crops during the growing period. Livestock may be of considerably more importance in terms of defining and cementing social relations (e.g. as bride-wealth) or as forms of investment than in terms of income⁴. The distinction between this system and the seminomadic pastoralism is one of degree in terms of relative mobility, allocation of labour to cultivation, and dependence on income from crops.

4 Examples are the Mangari (Horowitz, 1972), the Gogo of Tanzania (Rigby, 1969), the Berti of Sudan (Holy, 1974), and many of the Bangwaketse of Botswana (Gulbransen, 1980).

The three other groups of production systems, which are of less (but nevertheless of some) importance in terms of this study of water management, are the mixed farming systems in high-altitude areas⁵, the small-holder dairy systems, and ranching. These systems will occasionally be referred to later in this report.

5 This system is particularly important in highland Ethiopia (Cossins and Yemerou, 1974).